Why tail latency (p99) matters
WHAT IT TESTS: Tail-latency reasoning. OUTLINE: averages hide the worst experiences, and fan-out amplifies tails so most requests touch a slow path; causes include GC pauses, queueing, contention, and noisy neighbors.
Front-end performance budgets
WHAT IT TESTS: Performance governance. OUTLINE: a performance budget is an enforced limit on metrics like Core Web Vitals and bundle size, checked in CI to fail builds that regress.
Little's Law for capacity planning
WHAT IT TESTS: Queueing fundamentals. OUTLINE: L equals lambda times W, concurrency equals arrival rate times time in system; rearrange to size threads or concurrency for a target throughput and latency.
Capacity planning for annual traffic spikes
WHAT IT TESTS: Elastic capacity planning. OUTLINE: forecast peak from history, load-test to find per-unit capacity, use autoscaling with pre-warming and headroom, and protect with caching and graceful degradation.
Diagnose database CPU saturation under load
WHAT IT TESTS: DB performance diagnosis. OUTLINE: find the expensive queries via the database's stats, check for missing indexes and full scans, then fix with indexing, query rewrites, caching, or read replicas.
Load vs stress vs soak testing
WHAT IT TESTS: Performance-test vocabulary. OUTLINE: load tests expected traffic, stress pushes past limits to find the breaking point, soak runs sustained load for hours to expose leaks.
Golden signals for capacity planning
WHAT IT TESTS: Core observability fundamentals. OUTLINE: monitor the four golden signals, latency, traffic, errors, and saturation, from day one, watching percentiles and saturation to forecast scaling.
Canary a shared downstream microservice
WHAT IT TESTS: Mesh-based canary isolation. OUTLINE: route a slice of traffic to the canary via mesh rules, propagate context, and use distributed tracing to measure impact on upstream callers across the full path.
Auto-rollback on failed blue-green cutover
WHAT IT TESTS: Safe cutover design. OUTLINE: shift traffic gradually behind a smart router, use deep health checks plus real SLI monitoring, and auto-revert to blue on breach while blue stays warm.
Client-side vs server-side feature flags
WHAT IT TESTS: Flag evaluation placement. OUTLINE: client-side is fast and offline-capable but exposes flag logic and risks stale or leaked values; server-side keeps logic secret and consistent but adds latency.
Design automated canary analysis scoring
WHAT IT TESTS: Automated release gating. OUTLINE: track the golden-signal SLIs, compare canary to baseline statistically, weight and combine into a score with promote/rollback thresholds.
Diagnose a degraded canary release
WHAT IT TESTS: Canary decision-making. OUTLINE: check statistical significance versus baseline, confirm apples-to-apples comparison, isolate the cause via traces and logs, then weigh the regression against SLO budget.
Blue-green deploys with schema migrations
WHAT IT TESTS: Stateful deploy safety. OUTLINE: the shared database means both versions hit one schema, so breaking changes must be split into backward-compatible steps via expand-and-contract.
Measure ROI of toil reduction efforts
WHAT IT TESTS: Quantifying automation value. OUTLINE: track toil hours, percent of time on toil, incidents auto-resolved, and engineer cost saved, then frame as ROI and risk reduction. RED FLAG: vanity metrics or claiming success with no baseline measurement.
Design a centralized auto-remediation platform
WHAT IT TESTS: Self-healing system design. OUTLINE: event ingestion, a rules engine mapping alerts to playbooks, a sandboxed execution runtime, and guardrails like dry-run, rate limits, and rollback.
Design automated microservice provisioning workflow
WHAT IT TESTS: Golden-path platform design. OUTLINE: template scaffolding plus a pipeline that creates repo, CI/CD, and infra as code, with idempotent steps and rollback.
Systematically reduce noisy alert toil
WHAT IT TESTS: Data-driven toil reduction. OUTLINE: inventory alerts, measure frequency, actionability, and time cost, then prioritize by volume times effort. RED FLAG: muting alerts blindly or automating before measuring which ones actually hurt.
Idempotency in infrastructure provisioning scripts
WHAT IT TESTS: Safe-retry design. OUTLINE: idempotency means repeated runs converge to one end state; achieve it via desired-state reconciliation or idempotency keys with read-before-write.
Automate temporary elevated database access securely
WHAT IT TESTS: Just-in-time access design. OUTLINE: self-service request with approval, short-lived auto-expiring grants scoped to least privilege, and full audit logging.
Distinguishing toil from necessary ops work
WHAT IT TESTS: Defining toil precisely. OUTLINE: toil is manual, repetitive, automatable, tactical, and scales with growth; engineering and judgment-heavy work is not toil. RED FLAG: calling all ops work toil, or thinking toil means unimportant work.